Last year we reported our success in making a full-length cDNA clone of dengue virus type 2 (DEN2) from which infectious transcripts could be made. This year we have begun a collaboration to repeat this with another dengue serotype, DEN1. The work accomplished so far is described below. The left-hand 2.6 kb and right-hand 0.5 kb of DEN1 were amplified by RT-PCR using DEN1 virion RNA as template and a thermostable DNA polymerase with a low error rate. The primers used for the PCR incorporated an SP6 promoter just upstream of the DEN1 5' end and a unique Sac II site just downstream of the 3' end. These two cDNA fragments were cloned into the polylinker of the yeast shuttle vector pRS424, oriented properly. This DNA was linearized with an enzyme that cuts in the polylinker in between the two DEN cDNAs. Meanwhile, an 8 kb cDNA was made from DEN1 RNA by RT-PCR, using a more error-prone polymerase capable of long PCR. This "stuffer fragment" represents the middle of the DEN1 genome and overlaps both the left- and right-hand cDNAs contained in the pRS clone. The stuffer fragment and the linearized pRS DNA were used to cotransfect competent trp- yeast to TRP+. This selects for circular forms of the pRS plasmid, which carries theTRP1 gene necessary to confer the TRP+ phenotype. In yeast, double homologous recombination between the stuffer fragment and the left-hand and right-hand cDNAs creates a circular pRS plasmid containing the full-length DEN1 cDNA. TRP+ yeast colonies were screened by PCR to confirm the structure of the pRS plasmids they harbored. Plasmids apparently containing full-length DEN1 cDNA were transfected into E. coli, and large scale DNA preps were done. The structures of four of these E. coli derived plasmids were confirmed as correct by restriction analysis. Transcripts made from these four clones were electroporated into mammalian cells, but were not infectious. Subsequent analysis by transient expression using the vaccinia virus - T7 RNA polymerase system showed that these four clones did not make the expected pattern of DEN1 proteins. Sequencing of selected regions of two of these clones has been done, and shows that there are errors in the stuffer fragment sequences, including a stop codon in one case. Thus, our long PCR product may not be a faithful copy. We will continue to screen for more full-length clones to seek ones that might make infectious transcripts. Meanwhile, we plan to fully sequence one or more of the noninfectious clones, and then try to repair the mistakes, using cDNA made with the low error rate polymerase as the source of the cDNA for repair.